ElevenLabs launches real-time speech-to-speech models scoring 98.8% on Big Bench Audio
The Decoder · rss · 2026-10-21
ElevenLabs has released Eleven Realtime, a new family of speech-to-speech models built for real-time AI voice conversations, balancing speed and accuracy.
- The v1 Max variant scores 98.8% on the Big Bench Audio benchmark with a 0.57-second response time.
- According to ElevenLabs, no model on the Artificial Analysis leaderboard is both faster and more accurate.
- The models target low-latency, high-fidelity voice interaction for conversational AI applications.
More from Multimodal
- Stanford's LILIE loss cuts text-to-image gFD by 14 points over pixel-only training — StanfordUniversity · 2026-10-21
- Ex-Google Brain/DeepMind team launches Vyb, an AI design model with editable layered output — testingcatalog · 2026-10-21
- MOSS-Image's text encoder is multimodal and accepts image inputs — RandumbRedditor1000 · 2026-10-21
- Making a classroom case-explainer video with Opus 5.5 — technollama · 2026-10-21
- FastVideo releases FastH3 Omni weight collection with FP8 and NVFP4 ComfyUI builds — multimodalart · 2026-10-21
- Opus 5.5 distills a 30-min pi-durable lecture into a 7-min narrated explainer video — solyarisoftware · 2026-10-21